mirror of
https://github.com/explosion/spaCy.git
synced 2024-12-29 11:26:28 +03:00
db55577c45
* Remove unicode declarations * Remove Python 3.5 and 2.7 from CI * Don't require pathlib * Replace compat helpers * Remove OrderedDict * Use f-strings * Set Cython compiler language level * Fix typo * Re-add OrderedDict for Table * Update setup.cfg * Revert CONTRIBUTING.md * Revert lookups.md * Revert top-level.md * Small adjustments and docs [ci skip]
47 lines
1.7 KiB
Python
47 lines
1.7 KiB
Python
from ...symbols import ORTH, LEMMA, NORM
|
||
|
||
_exc = {}
|
||
|
||
_abbrev_exc = [
|
||
# Weekdays abbreviations
|
||
{ORTH: "дш", LEMMA: "дүшәмбе"},
|
||
{ORTH: "сш", LEMMA: "сишәмбе"},
|
||
{ORTH: "чш", LEMMA: "чәршәмбе"},
|
||
{ORTH: "пш", LEMMA: "пәнҗешәмбе"},
|
||
{ORTH: "җм", LEMMA: "җомга"},
|
||
{ORTH: "шб", LEMMA: "шимбә"},
|
||
{ORTH: "яш", LEMMA: "якшәмбе"},
|
||
# Months abbreviations
|
||
{ORTH: "гый", LEMMA: "гыйнвар"},
|
||
{ORTH: "фев", LEMMA: "февраль"},
|
||
{ORTH: "мар", LEMMA: "март"},
|
||
{ORTH: "мар", LEMMA: "март"},
|
||
{ORTH: "апр", LEMMA: "апрель"},
|
||
{ORTH: "июн", LEMMA: "июнь"},
|
||
{ORTH: "июл", LEMMA: "июль"},
|
||
{ORTH: "авг", LEMMA: "август"},
|
||
{ORTH: "сен", LEMMA: "сентябрь"},
|
||
{ORTH: "окт", LEMMA: "октябрь"},
|
||
{ORTH: "ноя", LEMMA: "ноябрь"},
|
||
{ORTH: "дек", LEMMA: "декабрь"},
|
||
# Number abbreviations
|
||
{ORTH: "млрд", LEMMA: "миллиард"},
|
||
{ORTH: "млн", LEMMA: "миллион"},
|
||
]
|
||
|
||
for abbr in _abbrev_exc:
|
||
for orth in (abbr[ORTH], abbr[ORTH].capitalize(), abbr[ORTH].upper()):
|
||
_exc[orth] = [{ORTH: orth, LEMMA: abbr[LEMMA], NORM: abbr[LEMMA]}]
|
||
_exc[orth + "."] = [{ORTH: orth + ".", LEMMA: abbr[LEMMA], NORM: abbr[LEMMA]}]
|
||
|
||
for exc_data in [ # "etc." abbreviations
|
||
{ORTH: "һ.б.ш.", NORM: "һәм башка шундыйлар"},
|
||
{ORTH: "һ.б.", NORM: "һәм башка"},
|
||
{ORTH: "б.э.к.", NORM: "безнең эрага кадәр"},
|
||
{ORTH: "б.э.", NORM: "безнең эра"},
|
||
]:
|
||
exc_data[LEMMA] = exc_data[NORM]
|
||
_exc[exc_data[ORTH]] = [exc_data]
|
||
|
||
TOKENIZER_EXCEPTIONS = _exc
|